The present invention relates, in general, to charge coupled devices and methods for the fabrication of the same. More specifically, the present invention is concerned with a structure of charge coupled electrode capable of improving charge transfer efficiency, and methods for fabricating the same.
A charge coupled device, which transfers pulse charges in one direction by use of the potential difference within a semiconductor device induced by the potential difference applied to each transfer electrode and which is widely utilized for a solid state image sensing device, and signal delay, is generally structured to have an array of fine transfer electrodes which are separated from one another by an insulating film on a silicon substrate.
For better understanding of the background of the present invention, a description of conventional charge coupled devices is to be given with reference to some drawings.
Referring to FIG. 1, there is shown a conventional structure of a charge coupled device. Such structure of charge coupled device is fabricated as follows.
First, into a p type silicon substrate 1 n type impurity ions are implanted, to form a buried charge coupled device (hereinafter referred to as "BCCD") region 2 over which an oxide film 3 serving as an insulating film is then formed entirely.
Over the oxide film 3, there is deposited a conductive layer of polysilicon, which is subsequently patterned by use of photoetch, so as to form a plurality of spaced-apart, parallel first transfer electrodes 13.
Thereafter, using the first transfer electrodes 13 as a mask, a barrier 9 is settled on the surface of the BCCD region 2 by ion implantation., and the first transfer electrodes 13 are insulated by an oxide film. Then, a plurality of spaced-apart, parallel second transfer electrodes 14 are formed of polysilicon between the first transfer electrodes.
In such charge coupled device fabricated, each of the first transfer electrodes 13 is paired with a nearby one of the second transfer electrodes 14. Each of these electrode pairs is connected with either a first clock pulse and a second clock pulse (H.PHI..sub.1, H.PHI..sub.2) which are alternatively applied to the electrode pairs.
Referring now to FIG. 2, there is illustrated the operating principle of the conventional two-phase charge coupled device. While FIG. 2A is an example of the first and the second clock pulses applied to the transfer electrode of the two-phase charge coupled device, FIG. 2B shows a potential distribution induced within a semiconductor when applying the first and the second clock pulses to the transfer electrodes, and a migration course of charges according to the potential distribution.
In detail, at t=1, the first clock pulse (H.PHI..sub.1) is in a state of "low", whereas the second clock pulse (H.PHI..sub.2) is in a state of "high". At the moment, a potential well becomes deepest at below the transfer electrode 13, so that the pulse charges are trapped ill this well.
At t=2, the first clock pulse (H.PHI..sub.1) is in a high state, whereas the second clock pulse (H.PHI..sub.2) is in a low state. Accordingly, the deepest potential well is formed below the first transfer electrode 13 applied with the first clock pulse (H.PHI..sub.1), and the potential well of the second transfer electrode 14 applied with the second clock pulse (H.PHI..sub.2) is risen, so that pulse charges move into below the first transfer 13 which has the deepest well and which is applied with the first clock pulse (H.PHI..sub.1).
At t=3, the pulse charges move like at t=2. In this point, the pulse charge has directivity to move only rightward by virtue of a potential barrier formed below the left electrode of the transfer electrode pair consisting of the first transfer electrode and the second transfer electrode.
Repetition of such first and second clock pulses (H.PHI..sub.1, H.PHI..sub.2) allows the pulse charge to be transferred.
However, the conventional charge coupled device is problematic in charge transfer. To scrutinize the potential distribution shown in FIG. 2B, it could be found that the edge portion below each the transfer electrodes shows a rapid change of the potential which allows the pulse charge to smoothly move thereat, whereas since the central portion shows equipotential distribution, the pulse charge moves into a neighboring electrode by not the force of electrical field but only diffusion by which the charge transfer is slow in velocity as well as difficult to complete.
Such phenomena are more apparent as the frequency applied to the transfer electrode increases. Therefore, the conventional charge coupled device operates in low performance at high frequencies.